Outline:¶

Introduction:¶

  • What is anomalous microwave emission? (AME)
  • Why is AKARI useful?

Our approach:¶

  • Compare Planck AME map with IR maps
  • Can we say more than "AME Correlates with dust"

Results:¶

  • All-sky comparison - AME vs. IRC
  • $\lambda$ Orionis -

Discussion:¶

  • Can we rule out spinning dust?
  • What about effect of uncertainties on the correlation?
  • Do we need better AME data?

Future (ongoing)) work:¶

  • Dust SED fittin (with help from Galliano-sensei) Tangential work:
  • Explore all-sky data with machine learning methods

Intro: What's in the microwave sky?¶

In [3]:
hp.mollview(lfi30ghz,
           title = 'Planck Low Frequency Instrument: 30 GHz All-sky map',
           norm = 'hist',
           cmap = 'rainbow',
           unit = '$K_{CMB}$')

alt text

From "Planck 2015 results. X. Diffuse component separation: Foreground maps":¶

https://arxiv.org/abs/1502.01588

Our approach¶

  • ### Compare AKARI, Planck, and IRAS maps to the Planck AME Component map:
  1. #### All-sky comparison
  2. #### Localized inspection of $\lambda$ Orionis
  • ### Major caveats:
    • #### Limited to ~1-degree resolution
    • #### No "ground truth" AME map !!
    • #### Envrionmental variations not well constrained on an all-sky basis

Intro: Why AKARI?¶

  • #### Dominated by PAH emission features
  • #### IRC 9 $\mu{}m$ to IRAS 12 $\mu{}m$ ratio may reveal trends in PAH ionization
In [7]:
from IPython.display import Image 
from IPython.core.display import HTML 

PATH = "../Plots" 
Image(filename = PATH + "/png/band-ratio-G100.png")
Out[7]:
  • $G_{0}$ indicates the interstellar radiation field relative to the solar neighborhood

All-sky results¶

Each IR band's intensity vs. AME Intensity:¶

In [22]:
#### 2D Kernel Density plots
In [23]:
Image('../Plots/png/AMEvsDust_allsky_allbands__mpsub_kde.png')
Out[23]:

What if scale by the radiation field strength ($U$)?¶

In [10]:
Image('../Plots/png/AMEvsDust_allsky_allbands__mpsub__UNorm_kde.png')
Out[10]:
  • $U$ is approximated as the dust radiance $R$ divided by the optical depth $\tau_{353 GHz}$
  • $R$ and $\tau_{353 GHz}$ are from the Planck PR1 thermal dust parameter maps

Spearman Correlation Matrix:¶

IR Bands' Intensity and AME

In [11]:
Image('../Plots/png/all_bands_corr_matrix_wAME_spearmanintensity.png')
Out[11]:

Spearman Correlations:¶

After normalizing data by ISRF ( ~ $\frac{R}{\tau_{353 GHz}}$)¶
In [12]:
Image('../Plots/png/all_bands_corr_matrix_wAME_spearmanU_norm.png')
Out[12]:

Almost no change when dividing by $U$¶

Spatial Variation of Correlation Strength¶

$I_{9}$ to $I_{AME}$ correlation strength for ~10 degree patches:¶

In [14]:
Image('../Plots/Allsky_Corr/Spearman_Map_nside8_A9toAME.png')
Out[14]:
  • ###### Looks like a Galaxy...
  • ###### Weaker correlation at higher latitudes

$I_{140}$ to $I_{AME}$ correlation strength for ~10 degree patches:¶

In [15]:
Image('../Plots/Allsky_Corr/Spearman_Map_nside8_A140toAME.png')
Out[15]:

What if we normalize by the dust radiance?¶

In [16]:
Image('../Plots/Allsky_Corr/RadNorm/Spearman_Map_nside8_A9toAME.png')
Out[16]:
  • Now the map flattens-out
  • Consistent with Hensley+ (2016)
  • ...but is this indicating lack of AME:PAH correlation?
  • Are we encountering too much noise/systematic effects in the maps?

$\lambda$ Orionis Results:¶

In [17]:
Image('../Plots/png/LOri_akari9_AMEcont_1dres.png')
Out[17]:
  • The binary star $\lambda$ Orionis, is located near the center of the ring.
In [19]:
Image('../Plots/lOrionis_grid_img.png')
Out[19]:

$\lambda$ Orionis Results:¶

Spearman Correlation Matrix

In [20]:
Image('../Plots/png/Allsky_corr_Lori.png')
Out[20]:
  • #### 18 micron to 65 micron bands correlate the worst?
  • #### AME apparently weaker at higher $T_{dust}$ pixels
  • #### Destruction of PAHs in the center? Something else?
  • #### ..stay tuned for dust SED modeling (with help from F. Galliano)

$\lambda$ Orionis Results:¶

What about PAH ionization fraction?¶

In [21]:
Image('../Plots/png/Lori_A9I12.png')
Out[21]:
  • Stronger AME relates to higher 9 to 12 micron intensity ratio
  • Interesting because both bands are weak in the central region
  • PAH ionization fraction needs more attention...

Conclusions:¶

All-sky:¶

  • #### All-sky dust comparisons are a mess (see F. Boulanger's talk)
  • #### Present data do not rule out spinning-PAHs, however:
  • #### They also do not indicate exclusive PAH-AME relationship

$\lambda$ Orionis¶

  • #### PAH-tracing emission correlates as well with AME as FIR in $\lambda$ Orionis
  • #### Correlation of 9 to 12 $\mu{}m$ ratio suggests PAH ionization needs more consideration.

Open questions:¶

  • What about noise and systematic affects?
    • Effect on correlation tests
  • If not PAHs, what about other spinning stuff? (nanosilicates, etc. see T. Hoang's talk)
    • "What about the magnetic dust?" - Magnetic dipole emission not dead yet?
  • Do we know enough about the dust in different environments? (F. Boulanger's talk again)
  • How well do the Planck Component maps really constrain AME?

Future (Ongoing) Work:¶

  • Dust SED modelling (with F. Galliano) -- taking consideration of uncertainties
  • Comparing Planck component separation with "blind" component separation techniques. (I.e. PCA, ICA, NMF)